idnits 2.17.1 draft-ietf-precis-7613bis-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- -- The abstract seems to indicate that this document obsoletes RFC3454, but the header doesn't have an 'Obsoletes:' line to match this. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (May 1, 2017) is 2552 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) ** Obsolete normative reference: RFC 7564 (Obsoleted by RFC 8264) -- Possible downref: Non-RFC (?) normative reference: ref. 'UAX11' -- Possible downref: Non-RFC (?) normative reference: ref. 'Unicode' -- Obsolete informational reference (is this intentional?): RFC 4013 (ref. 'Err1812') (Obsoleted by RFC 7613) -- Obsolete informational reference (is this intentional?): RFC 3454 (Obsoleted by RFC 7564) -- Obsolete informational reference (is this intentional?): RFC 3501 (Obsoleted by RFC 9051) -- Duplicate reference: RFC4013, mentioned in 'RFC4013', was also mentioned in 'Err1812'. -- Obsolete informational reference (is this intentional?): RFC 4013 (Obsoleted by RFC 7613) -- Obsolete informational reference (is this intentional?): RFC 7613 (Obsoleted by RFC 8265) Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 10 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group P. Saint-Andre 3 Internet-Draft Filament 4 Obsoletes: 7613 (if approved) A. Melnikov 5 Intended status: Standards Track Isode Ltd 6 Expires: November 2, 2017 May 1, 2017 8 Preparation, Enforcement, and Comparison of Internationalized Strings 9 Representing Usernames and Passwords 10 draft-ietf-precis-7613bis-07 12 Abstract 14 This document describes updated methods for handling Unicode strings 15 representing usernames and passwords. The previous approach was 16 known as SASLprep (RFC 4013) and was based on stringprep (RFC 3454). 17 The methods specified in this document provide a more sustainable 18 approach to the handling of internationalized usernames and 19 passwords. The preparation, enforcement, and comparison of 20 internationalized strings (PRECIS) framework, RFC 7564, obsoletes RFC 21 3454, and this document obsoletes RFC 7613. 23 Status of This Memo 25 This Internet-Draft is submitted in full conformance with the 26 provisions of BCP 78 and BCP 79. 28 Internet-Drafts are working documents of the Internet Engineering 29 Task Force (IETF). Note that other groups may also distribute 30 working documents as Internet-Drafts. The list of current Internet- 31 Drafts is at http://datatracker.ietf.org/drafts/current/. 33 Internet-Drafts are draft documents valid for a maximum of six months 34 and may be updated, replaced, or obsoleted by other documents at any 35 time. It is inappropriate to use Internet-Drafts as reference 36 material or to cite them other than as "work in progress." 38 This Internet-Draft will expire on November 2, 2017. 40 Copyright Notice 42 Copyright (c) 2017 IETF Trust and the persons identified as the 43 document authors. All rights reserved. 45 This document is subject to BCP 78 and the IETF Trust's Legal 46 Provisions Relating to IETF Documents 47 (http://trustee.ietf.org/license-info) in effect on the date of 48 publication of this document. Please review these documents 49 carefully, as they describe your rights and restrictions with respect 50 to this document. Code Components extracted from this document must 51 include Simplified BSD License text as described in Section 4.e of 52 the Trust Legal Provisions and are provided without warranty as 53 described in the Simplified BSD License. 55 Table of Contents 57 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 58 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4 59 3. Usernames . . . . . . . . . . . . . . . . . . . . . . . . . . 5 60 3.1. Definition . . . . . . . . . . . . . . . . . . . . . . . 5 61 3.2. Case Mapping vs. Case Preservation . . . . . . . . . . . 6 62 3.3. UsernameCaseMapped Profile . . . . . . . . . . . . . . . 7 63 3.3.1. Rules . . . . . . . . . . . . . . . . . . . . . . . . 7 64 3.3.2. Preparation . . . . . . . . . . . . . . . . . . . . . 8 65 3.3.3. Enforcement . . . . . . . . . . . . . . . . . . . . . 8 66 3.3.4. Comparison . . . . . . . . . . . . . . . . . . . . . 8 67 3.4. UsernameCasePreserved Profile . . . . . . . . . . . . . . 9 68 3.4.1. Rules . . . . . . . . . . . . . . . . . . . . . . . . 9 69 3.4.2. Preparation . . . . . . . . . . . . . . . . . . . . . 9 70 3.4.3. Enforcement . . . . . . . . . . . . . . . . . . . . . 9 71 3.4.4. Comparison . . . . . . . . . . . . . . . . . . . . . 10 72 3.5. Application-Layer Constructs . . . . . . . . . . . . . . 10 73 3.6. Examples . . . . . . . . . . . . . . . . . . . . . . . . 10 74 4. Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . 12 75 4.1. Definition . . . . . . . . . . . . . . . . . . . . . . . 12 76 4.2. OpaqueString Profile . . . . . . . . . . . . . . . . . . 13 77 4.2.1. Preparation . . . . . . . . . . . . . . . . . . . . . 13 78 4.2.2. Enforcement . . . . . . . . . . . . . . . . . . . . . 14 79 4.2.3. Comparison . . . . . . . . . . . . . . . . . . . . . 14 80 4.3. Examples . . . . . . . . . . . . . . . . . . . . . . . . 15 81 5. Use in Application Protocols . . . . . . . . . . . . . . . . 16 82 6. Migration . . . . . . . . . . . . . . . . . . . . . . . . . . 16 83 6.1. Usernames . . . . . . . . . . . . . . . . . . . . . . . . 16 84 6.2. Passwords . . . . . . . . . . . . . . . . . . . . . . . . 18 85 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 19 86 7.1. UsernameCaseMapped Profile . . . . . . . . . . . . . . . 19 87 7.2. UsernameCasePreserved Profile . . . . . . . . . . . . . . 19 88 7.3. OpaqueString Profile . . . . . . . . . . . . . . . . . . 20 89 7.4. Stringprep Profile . . . . . . . . . . . . . . . . . . . 21 90 8. Security Considerations . . . . . . . . . . . . . . . . . . . 21 91 8.1. Password/Passphrase Strength . . . . . . . . . . . . . . 21 92 8.2. Identifier Comparison . . . . . . . . . . . . . . . . . . 21 93 8.3. Reuse of PRECIS . . . . . . . . . . . . . . . . . . . . . 21 94 8.4. Reuse of Unicode . . . . . . . . . . . . . . . . . . . . 21 95 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 21 96 9.1. Normative References . . . . . . . . . . . . . . . . . . 21 97 9.2. Informative References . . . . . . . . . . . . . . . . . 22 98 Appendix A. Changes from RFC 7613 . . . . . . . . . . . . . . . 24 99 Appendix B. Acknowledgements . . . . . . . . . . . . . . . . . . 24 100 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 25 102 1. Introduction 104 Usernames and passwords are widely used for authentication and 105 authorization on the Internet, either directly when provided in 106 plaintext (as in the PLAIN Simple Authentication and Security Layer 107 (SASL) mechanism [RFC4616] and the HTTP Basic scheme [RFC7617]) or 108 indirectly when provided as the input to a cryptographic algorithm 109 such as a hash function (as in the Salted Challenge Response 110 Authentication Mechanism (SCRAM) SASL mechanism [RFC5802] and the 111 HTTP Digest scheme [RFC7616]). 113 To increase the likelihood that the input and comparison of usernames 114 and passwords will work in ways that make sense for typical users 115 throughout the world, this document defines rules for preparing, 116 enforcing, and comparing internationalized strings that represent 117 usernames and passwords. Such strings consist of code points from 118 the Unicode coded character set [Unicode], with special attention to 119 code points outside the ASCII range [RFC20]. The rules for handling 120 such strings are specified through profiles of the string classes 121 defined in the preparation, enforcement, and comparison of 122 internationalized strings (PRECIS) framework specification [RFC7564]. 124 Profiles of the PRECIS framework enable software to handle Unicode 125 code points outside the ASCII range in an automated way, so that such 126 code points are treated carefully and consistently in application 127 protocols. In large measure, these profiles are designed to protect 128 application developers from the potentially negative consequences of 129 supporting the full range of Unicode code points. For instance, in 130 almost all application protocols it would be dangerous to treat the 131 Unicode code point SUPERSCRIPT ONE (U+00B9) as equivalent to DIGIT 132 ONE (U+0031), because that would result in false positives during 133 comparison, authentication, and authorization (e.g., an attacker 134 could easy spoof an account "user1@example.com"). 136 Whereas a naive use of Unicode would make such attacks trivially 137 easy, the PRECIS profile defined here for usernames generally 138 protects applications from inadvertently causing such problems. 139 (Similar considerations apply to passwords, although here it is 140 desirable to support a wider range of characters so as to maximize 141 entropy for purposes of authentication.) 143 The methods defined here might be applicable wherever usernames or 144 passwords are used. However, the methods are not intended for use in 145 preparing strings that are not usernames (e.g., Lightweight Directory 146 Access Protocol (LDAP) distinguished names), nor in cases where 147 identifiers or secrets are not strings (e.g., keys and certificates) 148 or require specialized handling. 150 This document obsoletes RFC 4013 (the SASLprep profile of stringprep 151 [RFC3454]) but can be used by technologies other than SASL [RFC4422], 152 such as HTTP authentication as specified in [RFC7617] and [RFC7616]. 154 This document does not modify the handling of internationalized 155 strings in usernames and passwords as prescribed by existing 156 application protocols that use SASLprep. If the community that uses 157 such an application protocol wishes to modernize its handling of 158 internationalized strings to use PRECIS instead of stringprep, it 159 needs to explicitly update the existing application protocol 160 definition (one example is [RFC7622]. Non-coordinated updates to 161 protocol implementations are discouraged because they can have a 162 negative impact on interoperability and security. 164 2. Terminology 166 A "username" or "user identifier" is a string of characters 167 designating an account on a computing device or system, often but not 168 necessarily for use by a person. Although some devices and system 169 might allow a username to be part or all of a person's name, and a 170 person might want their account designator to be part or all of their 171 name, because of the complexities involved that outcome is not 172 guaranteed for all human names on all computing devices or systems 173 that follow the rules defined in this specification. Protocol 174 designers and application developers who wish to allow a wider range 175 of characters are encouraged to consider a separation between more 176 restrictive account identifiers and more expressive display names. 178 A "password" is a string of characters that allows access to a 179 computing device or system, often associated with a particular 180 username. A password is not literally limited to a word, because a 181 password could be a passphrase consisting of more than one word, 182 perhaps separated by spaces, punctuation, or other non-alphanumeric 183 characters. 185 Some SASL mechanisms (e.g., CRAM-MD5, DIGEST-MD5, and SCRAM) specify 186 that the authentication identity used in the context of such 187 mechanisms is a "simple user name" (see Section 2 of [RFC4422] as 188 well as [RFC4013]). Various application technologies also assume 189 that the identity of a user or account takes the form of a username 190 (e.g., authentication for the Hypertext Transfer Protocol as 191 specified in [RFC7617] and [RFC7616]), whether or not they use SASL. 192 Note well that the exact form of a username in any particular SASL 193 mechanism or application technology is a matter for implementation 194 and deployment, and that a username does not necessarily map to any 195 particular application identifier. 197 Many important terms used in this document are defined in [RFC5890], 198 [RFC6365], [RFC7564], and [Unicode]. The term "non-ASCII space" 199 refers to any Unicode code point having a Unicode general category of 200 "Zs", with the exception of U+0020 (here called "ASCII space"). 202 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 203 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 204 "OPTIONAL" in this document are to be interpreted as described in 205 [RFC2119]. 207 3. Usernames 209 3.1. Definition 211 This document specifies that a username is a string of Unicode code 212 points [Unicode] that is structured as an ordered sequence of 213 "userparts" and expressed in a standard Unicode Encoding Form (such 214 as UTF-8 [RFC3629]). A userpart is allowed to contain only code 215 points that are allowed by the PRECIS IdentifierClass defined in 216 Section 4.2 of [RFC7564], and thus consists almost exclusively of 217 letters and digits. A username can consist of a single userpart or a 218 space-separated sequence of userparts. 220 The syntax for a username is defined as follows, using the Augmented 221 Backus-Naur Form (ABNF) [RFC5234]. 223 username = userpart *(1*SP userpart) 224 userpart = 1*(idpoint) 225 ; 226 ; an "idpoint" is a Unicode code point that 227 ; can be contained in a string conforming to 228 ; the PRECIS IdentifierClass 229 ; 231 All code points and blocks not explicitly allowed in the PRECIS 232 IdentifierClass are disallowed; this includes private use code 233 points, surrogate code points, and the other code points and blocks 234 that were defined as "Prohibited Output" in [RFC4013]. In addition, 235 common constructions such as "user@example.com" (e.g., the Network 236 Access Identifier from [RFC7542]) are allowed as usernames under this 237 specification, as they were under [RFC4013]. 239 Implementation Note: The username construct defined in this 240 document does not necessarily match what all deployed applications 241 might refer to as a "username" or "userid" but instead provides a 242 relatively safe subset of Unicode code points that can be used in 243 existing SASL mechanisms and in application protocols that use 244 SASL, and even in most application protocols that do not currently 245 use SASL. 247 A username MUST NOT be zero bytes in length. This rule is to be 248 enforced after any normalization and mapping of code points. 250 In protocols that provide usernames as input to a cryptographic 251 algorithm such as a hash function, the client will need to perform 252 enforcement of the rules for the UsernameCaseMapped or 253 UsernameCasePreserved profile before applying the algorithm. 255 This specification defines two profiles for usernames: one that 256 performs case mapping and one that performs case preservation (see 257 further discussion under Section 3.2). 259 3.2. Case Mapping vs. Case Preservation 261 In order to accommodate the widest range of username constructs in 262 applications, this document defines two username profiles: 263 UsernameCaseMapped and UsernameCasePreserved. These two profiles 264 differ only in the Case-Mapping Rule and are otherwise identical. 266 Case mapping is a matter for the application protocol, protocol 267 implementation, or end deployment. In general, this document 268 suggests that it is preferable to apply the UsernameCaseMapped 269 profile and therefore perform case mapping, because not doing so can 270 lead to false positives during authentication and authorization (as 271 described in [RFC6943]) and can result in confusion among end users, 272 given the prevalence of case mapping in many existing protocols and 273 applications. However, there can be good reasons to apply the 274 UsernameCasePreserved profile and thus not perform case mapping, such 275 as backward compatibility with deployed infrastructure. 277 In particular: 279 o SASL mechanisms that follow the recommendations in this document 280 MUST specify whether and when case mapping is to be applied to 281 authentication identifiers. SASL mechanisms SHOULD delay any case 282 mapping to the last possible moment, such as when doing a lookup 283 by username, performing username comparisons, or generating a 284 cryptographic salt from a username (if the last possible moment 285 happens on the server, then decisions about case mapping can be a 286 matter of deployment policy). In keeping with [RFC4422], SASL 287 mechanisms are not to apply this or any other profile to 288 authorization identifiers, only to authentication identifiers. 290 o Application protocols that use SASL (such as IMAP [RFC3501] and 291 the Extensible Messaging and Presence Protocol (XMPP) [RFC6120]) 292 and that directly reuse this profile MUST specify whether or not 293 case mapping is to be applied to authorization identifiers. Such 294 "SASL application protocols" SHOULD delay any case-mapping of 295 authorization identifiers to the last possible moment, which 296 happens to necessarily be on the server side (this enables 297 decisions about case mapping to be a matter of deployment policy). 298 In keeping with [RFC4422], SASL application protocols are not to 299 apply this or any other profile to authentication identifiers, 300 only to authorization identifiers. 302 o Application protocols that do not use SASL (such as HTTP 303 authentication with the HTTP Basic and Digest schemes as specified 304 in [RFC7617] and [RFC7616]) but that directly reuse this profile 305 MUST specify whether and when case mapping is to be applied to 306 authentication identifiers or authorization identifiers, or both. 307 Such "non-SASL application protocols" SHOULD delay any case 308 mapping to the last possible moment, such as when doing a lookup 309 by username, performing username comparisons, or generating a 310 cryptographic salt from a username (if the last possible moment 311 happens on the server, then decisions about case mapping can be a 312 matter of deployment policy). 314 If the specification for a SASL mechanism, SASL application protocol, 315 or non-SASL application protocol uses the UsernameCaseMapped profile, 316 it MUST clearly describe whether case mapping is to be applied at the 317 level of the protocol itself, implementations thereof, or service 318 deployments (each of these approaches can be legitimate, depending on 319 the application in question). 321 3.3. UsernameCaseMapped Profile 323 3.3.1. Rules 325 The following rules are defined for use within the UsernameCaseMapped 326 profile of the PRECIS IdentifierClass. 328 1. Width-Mapping Rule: Map fullwidth and halfwidth code points to 329 their decomposition mappings (see Unicode Standard Annex #11 330 [UAX11]). 332 2. Additional Mapping Rule: There is no additional mapping rule. 334 3. Case-Mapping Rule: Map uppercase and titlecase code points to 335 their lowercase equivalents, preferably using the Unicode 336 toLower() operation as defined in the Unicode Standard [Unicode]; 337 see further discussion in Section 3.2. 339 4. Normalization Rule: Apply Unicode Normalization Form C (NFC) to 340 all strings. 342 5. Directionality Rule: Apply the "Bidi Rule" defined in [RFC5893] 343 to strings that contain right-to-left code points (i.e., each of 344 the six conditions of the Bidi Rule must be satisfied); for 345 strings that do not contain right-to-left code points, there is 346 no special processing for directionality. 348 3.3.2. Preparation 350 An entity that prepares a string for subsequent enforcement according 351 to this profile MUST proceed as follows (applying the steps in the 352 order shown). 354 1. Apply the width-mapping rule specified in Section 3.3.1. It is 355 necessary to apply the rule at this point because otherwise the 356 PRECIS "HasCompat" category specified in Section 9.17 of 357 [RFC7564] would forbid fullwidth and halfwidth code points. 359 2. Ensure that the string consists only of Unicode code points that 360 are explicitly allowed by the PRECIS IdentifierClass defined in 361 Section 4.2 of [RFC7564]. 363 3.3.3. Enforcement 365 An entity that performs enforcement according to this profile MUST 366 prepare a string as described in Section 3.3.2 and MUST also apply 367 the following rules specified in Section 3.3.1 in the order shown: 369 1. Case-Mapping Rule 371 2. Normalization Rule 373 3. Directionality Rule 375 After all of the foregoing rules have been enforced, the entity MUST 376 ensure that the username is not zero bytes in length (this is done 377 after enforcing the rules to prevent applications from mistakenly 378 omitting a username entirely, because when internationalized strings 379 are accepted, a non-empty sequence of characters can result in a 380 zero-length username after canonicalization). 382 3.3.4. Comparison 384 An entity that performs comparison of two strings according to this 385 profile MUST prepare each string as specified in Section 3.3.2 and 386 then MUST enforce the rules specified in Section 3.3.3. The two 387 strings are to be considered equivalent if and only if they are an 388 exact octet-for-octet match (sometimes called "bit-string identity"). 390 3.4. UsernameCasePreserved Profile 392 3.4.1. Rules 394 The following rules are defined for use within the 395 UsernameCasePreserved profile of the PRECIS IdentifierClass. 397 1. Width-Mapping Rule: Map fullwidth and halfwidth code points to 398 their decomposition mappings (see Unicode Standard Annex #11 399 [UAX11]). 401 2. Additional Mapping Rule: There is no additional mapping rule. 403 3. Case-Mapping Rule: There is no case-mapping rule. 405 4. Normalization Rule: Apply Unicode Normalization Form C (NFC) to 406 all strings. 408 5. Directionality Rule: Apply the "Bidi Rule" defined in [RFC5893] 409 to strings that contain right-to-left code points (i.e., each of 410 the six conditions of the Bidi Rule must be satisfied); for 411 strings that do not contain right-to-left code points, there is 412 no special processing for directionality. 414 3.4.2. Preparation 416 An entity that prepares a string for subsequent enforcement according 417 to this profile MUST proceed as follows (applying the steps in the 418 order shown). 420 1. Apply the width-mapping rule specified in Section 3.3.1. It is 421 necessary to apply the rule at this point because otherwise the 422 PRECIS "HasCompat" category specified in Section 9.17 of 423 [RFC7564] would forbid fullwidth and halfwidth code points. 425 2. Ensure that the string consists only of Unicode code points that 426 are explicitly allowed by the PRECIS IdentifierClass defined in 427 Section 4.2 of [RFC7564]. 429 3.4.3. Enforcement 431 An entity that performs enforcement according to this profile MUST 432 prepare a string as described in Section 3.4.2 and MUST also apply 433 the following rules specified in Section 3.4.1 in the order shown: 435 1. Normalization Rule 437 2. Directionality Rule 439 After all of the foregoing rules have been enforced, the entity MUST 440 ensure that the username is not zero bytes in length (this is done 441 after enforcing the rules to prevent applications from mistakenly 442 omitting a username entirely, because when internationalized strings 443 are accepted, a non-empty sequence of characters can result in a 444 zero-length username after canonicalization). 446 3.4.4. Comparison 448 An entity that performs comparison of two strings according to this 449 profile MUST prepare each string as specified in Section 3.4.2 and 450 then MUST enforce the rules specified in Section 3.4.3. The two 451 strings are to be considered equivalent if and only if they are an 452 exact octet-for-octet match (sometimes called "bit-string identity"). 454 3.5. Application-Layer Constructs 456 Both the UsernameCaseMapped and UsernameCasePreserved profiles enable 457 an application protocol, implementation, or deployment to create 458 application-layer constructs such as a username that is a space- 459 separated set of userparts like "Firstname Middlename Lastname". 460 Although such a construct is not a profile of the PRECIS 461 IdentifierClass (because U+0020 SPACE is not allowed in the 462 IdentifierClass), it can be created at the application layer because 463 U+0020 SPACE can be used as a separator between instances of the 464 PRECIS IdentifierClass (e.g., userparts as defined in this 465 specification). 467 3.6. Examples 469 The following examples illustrate a small number of userparts (not 470 usernames) that are consistent with the format defined above (note 471 that the characters "<" and ">" are used here to delineate the actual 472 userparts and are not part of the userpart strings). 474 +--------------------------+---------------------------------+ 475 | # | Userpart | Notes | 476 +--------------------------+---------------------------------+ 477 | 1 | | The at-sign is allowed in the | 478 | | | PRECIS IdentifierClass | 479 +--------------------------+---------------------------------+ 480 | 2 | | | 481 +--------------------------+---------------------------------+ 482 | 3 | | The third character is LATIN | 483 | | | SMALL LETTER SHARP S (U+00DF) | 484 +--------------------------+---------------------------------+ 485 | 4 | <π> | A userpart of GREEK SMALL | 486 | | | LETTER PI (U+03C0) | 487 +--------------------------+---------------------------------+ 488 | 5 | <Σ> | A userpart of GREEK CAPITAL | 489 | | | LETTER SIGMA (U+03A3) | 490 +--------------------------+---------------------------------+ 491 | 6 | <σ> | A userpart of GREEK SMALL | 492 | | | LETTER SIGMA (U+03C3) | 493 +--------------------------+---------------------------------+ 494 | 7 | <ς> | A userpart of GREEK SMALL | 495 | | | LETTER FINAL SIGMA (U+03C2) | 496 +--------------------------+---------------------------------+ 498 Table 1: A Sample of Legal Userparts 500 Several points are worth noting. Regarding examples 2 and 3: 501 although in German the character eszett (LATIN SMALL LETTER SHARP S 502 (U+00DF)) can mostly be used interchangeably with the two characters 503 "ss", the userparts in these examples are different and (if desired) 504 a server would need to enforce a registration policy that disallows 505 one of them if the other is registered. Regarding examples 5, 6, and 506 7: optional case-mapping of GREEK CAPITAL LETTER SIGMA (U+03A3) to 507 lowercase (i.e., to GREEK SMALL LETTER SIGMA (U+03C3)) during 508 comparison would result in matching the userparts in examples 5 and 509 6; however, because the PRECIS mapping rules do not account for the 510 special status of GREEK SMALL LETTER FINAL SIGMA (U+03C2), the 511 userparts in examples 5 and 7 or examples 6 and 7 would not be 512 matched during comparison. 514 The following examples illustrate strings that are not valid 515 userparts (not usernames) because they violate the format defined 516 above. 518 +--------------------------+---------------------------------+ 519 | # | Non-Userpart String | Notes | 520 +--------------------------+---------------------------------+ 521 | 8 | | Space (U+0020) is disallowed in | 522 | | | the userpart | 523 +--------------------------+---------------------------------+ 524 | 9 | <> | Zero-length userpart | 525 +--------------------------+---------------------------------+ 526 | 10| | The sixth character is ROMAN | 527 | | | NUMERAL FOUR (U+2163) | 528 +--------------------------+---------------------------------+ 529 | 11| <♚> | A user part of BLACK CHESS KING | 530 | | | (U+265A) | 531 +--------------------------+---------------------------------+ 533 Table 2: A Sample of Strings That Violate the Userpart Rule 535 Here again, several points are worth noting. Regarding example 8: 536 although this is not a valid userpart, it is a valid username because 537 it is a space-separated sequence of userparts. Regarding example 10: 538 the Unicode code point ROMAN NUMERAL FOUR (U+2163) has a 539 compatibility equivalent of the string formed of LATIN CAPITAL LETTER 540 I (U+0049) and LATIN CAPITAL LETTER V (U+0056), but code points with 541 compatibility equivalents are not allowed in the PRECIS 542 IdentifierClass. Regarding example 11: symbol characters such as 543 BLACK CHESS KING (U+265A) are not allowed in the PRECIS 544 IdentifierClass. 546 4. Passwords 548 4.1. Definition 550 This document specifies that a password is a string of Unicode code 551 points [Unicode] that is conformant to the OpaqueString profile 552 (specified below) of the PRECIS FreeformClass defined in Section 4.3 553 of [RFC7564], and that is expressed in a standard Unicode Encoding 554 Form (such as UTF-8 [RFC3629]). 556 The syntax for a password is defined as follows, using the Augmented 557 Backus-Naur Form (ABNF) [RFC5234]. 559 password = 1*(freepoint) 560 ; 561 ; a "freepoint" is a Unicode code point that 562 ; can be contained in a string conforming to 563 ; the PRECIS FreeformClass 564 ; 566 All code points and blocks not explicitly allowed in the PRECIS 567 FreeformClass are disallowed; this includes private use code points, 568 surrogate code points, and the other code points and blocks defined 569 as "Prohibited Output" in Section 2.3 of RFC 4013 (when corrected per 570 [Err1812]). 572 A password MUST NOT be zero bytes in length. This rule is to be 573 enforced after any normalization and mapping of code points. 575 Note: Some existing systems allow an empty string in places where 576 a password would be expected (e.g., command-line tools that might 577 be called from an automated script, or servers that might need to 578 be restarted without human intervention). From the perspective of 579 this document (and RFC 4013 before it), these empty strings are 580 not passwords but are workarounds for the practical difficulty of 581 using passwords in certain scenarios. The prohibition of zero- 582 length passwords is not a recommendation regarding password 583 strength (because a password of only one byte is highly insecure) 584 but is meant to prevent applications from mistakenly omitting a 585 password entirely; such an outcome is possible when 586 internationalized strings are accepted, because a non-empty 587 sequence of characters can result in a zero-length password after 588 canonicalization. 590 In protocols that provide passwords as input to a cryptographic 591 algorithm such as a hash function, the client will need to perform 592 enforcement of the rules for the OpaqueString profile before applying 593 the algorithm, because the password is not available to the server in 594 plaintext form. 596 4.2. OpaqueString Profile 598 The definition of the OpaqueString profile is provided in the 599 following sections, including detailed information about preparation, 600 enforcement, and comparison (for details on the distinction between 601 these actions, refer to [RFC7564]). 603 4.2.1. Preparation 605 An entity that prepares a string according to this profile MUST 606 ensure that the string consists only of Unicode code points that are 607 explicitly allowed by the FreeformClass base string class defined in 608 [RFC7564]. 610 4.2.2. Enforcement 612 An entity that performs enforcement according to this profile MUST 613 prepare a string as described in Section 4.2.1 and MUST also apply 614 the rules specified below for the OpaqueString profile (these rules 615 MUST be applied in the order shown): 617 1. Width-Mapping Rule: Fullwidth and halfwidth code points MUST NOT 618 be mapped to their decomposition mappings (see Unicode Standard 619 Annex #11 [UAX11]). 621 2. Additional Mapping Rule: Any instances of non-ASCII space MUST be 622 mapped to ASCII space (U+0020); a non-ASCII space is any Unicode 623 code point having a Unicode general category of "Zs" (with the 624 exception of U+0020). As was the case in RFC 4013, the inclusion 625 of only ASCII space prevents confusion with various non-ASCII 626 space code points, many of which are difficult to reproduce 627 across different input methods. 629 3. Case-Mapping Rule: There is no case mapping rule (because mapping 630 uppercase and titlecase code points to their lowercase 631 equivalents would lead to false positives and thus to reduced 632 security). 634 4. Normalization Rule: Unicode Normalization Form C (NFC) MUST be 635 applied to all strings. 637 5. Directionality Rule: There is no directionality rule. The "Bidi 638 Rule" (defined in [RFC5893]) and similar rules are unnecessary 639 and inapplicable to passwords, because they can reduce the 640 repertoire of characters that are allowed in a string and 641 therefore reduce the amount of entropy that is possible in a 642 password. Such rules are intended to minimize the possibility 643 that the same string will be displayed differently on a layout 644 system set for right-to-left display and a layout system set for 645 left-to-right display; however, passwords are typically not 646 displayed at all and are rarely meant to be interoperable across 647 different layout systems in the way that non-secret strings like 648 domain names and usernames are. Furthermore, it is perfectly 649 acceptable for opaque strings other than passwords to be 650 presented differently in different layout systems, as long as the 651 presentation is consistent in any given layout system. 653 4.2.3. Comparison 655 An entity that performs comparison of two strings according to this 656 profile MUST prepare each string as specified in Section 4.2.1 and 657 then MUST enforce the rules specified in Section 4.2.2. The two 658 strings are to be considered equivalent if and only if they are an 659 exact octet-for-octet match (sometimes called "bit-string identity"). 661 4.3. Examples 663 The following examples illustrate a small number of passwords that 664 are consistent with the format defined above (note that the 665 characters "<" and ">" are used here to delineate the actual 666 passwords and are not part of the password strings). 668 +------------------------------------+------------------------------+ 669 | # | Password | Notes | 670 +------------------------------------+------------------------------+ 671 | 12| | ASCII space is allowed | 672 +------------------------------------+------------------------------+ 673 | 13| | Differs by case from | 674 | | | example 12 | 675 +------------------------------------+------------------------------+ 676 | 14| <πßå> | Non-ASCII letters are OK | 677 | | | (e.g., GREEK SMALL LETTER | 678 | | | PI (U+03C0)) | 679 +------------------------------------+------------------------------+ 680 | 15| | Symbols are OK (e.g., BLACK | 681 | | | DIAMOND SUIT (U+2666)) | 682 +------------------------------------+------------------------------+ 683 | 16| | OGHAM SPACE MARK (U+1680) is | 684 | | | mapped to U+0020, and thus | 685 | | | the full string is mapped to | 686 | | | | 687 +------------------------------------+------------------------------+ 689 Table 3: A Sample of Legal Passwords 691 The following example illustrates a string that is not a valid 692 password because it violates the format defined above. 694 +------------------------------------+------------------------------+ 695 | # | Password | Notes | 696 +------------------------------------+------------------------------+ 697 | 17| | Controls are disallowed | 698 +------------------------------------+------------------------------+ 700 Table 4: A String That Violates the Password Rules 702 5. Use in Application Protocols 704 This specification defines only the PRECIS-based rules for the 705 handling of strings conforming to the UsernameCaseMapped and 706 UsernameCasePreserved profiles of the PRECIS IdentifierClass, and 707 strings conforming to the OpaqueString profile of the PRECIS 708 FreeformClass. It is the responsibility of an application protocol 709 to specify the protocol slots in which such strings can appear, the 710 entities that are expected to enforce the rules governing such 711 strings, and at what points during protocol processing or interface 712 handling the rules need to be enforced. See Section 6 of [RFC7564] 713 for guidelines on using PRECIS profiles in applications. 715 Above and beyond the PRECIS-based rules specified here, application 716 protocols can also define application-specific rules governing such 717 strings (rules regarding minimum or maximum length, further 718 restrictions on allowable code points or character ranges, safeguards 719 to mitigate the effects of visually similar characters, etc.), 720 application-layer constructs (see Section 3.5), and related matters. 722 Some PRECIS profile definitions encourage entities that enforce the 723 rules to be liberal in what they accept. However, for usernames and 724 passwords such a policy can be problematic, because it can lead to 725 false positives. An in-depth discussion can be found in [RFC6943]. 727 Applying the rules for any given PRECIS profile is not necessarily an 728 idempotent procedure for all code points. Therefore, implementations 729 might need to apply the rules more than once to an internationalized 730 string. 732 6. Migration 734 The rules defined in this specification differ slightly from those 735 defined by the SASLprep specification [RFC4013]. The following 736 sections describe these differences, along with their implications 737 for migration, in more detail. 739 6.1. Usernames 741 Deployments that currently use SASLprep for handling usernames might 742 need to scrub existing data when they migrate to the rules defined in 743 this specification. In particular: 745 o SASLprep specified the use of Unicode Normalization Form KC 746 (NFKC), whereas the UsernameCaseMapped and UsernameCasePreserved 747 profiles employ Unicode Normalization Form C (NFC). In practice, 748 this change is unlikely to cause significant problems, because 749 NFKC provides methods for mapping Unicode code points with 750 compatibility equivalents to those equivalents, whereas the PRECIS 751 IdentifierClass entirely disallows Unicode code points with 752 compatibility equivalents (i.e., during comparison, NFKC is more 753 "aggressive" about finding matches than NFC). A few examples 754 might suffice to indicate the nature of the problem: 756 1. LATIN SMALL LETTER LONG S (U+017F) is compatibility equivalent 757 to LATIN SMALL LETTER S (U+0073). 759 2. ROMAN NUMERAL FOUR (U+2163) is compatibility equivalent to 760 LATIN CAPITAL LETTER I (U+0049) and LATIN CAPITAL LETTER V 761 (U+0056). 763 3. LATIN SMALL LIGATURE FI (U+FB01) is compatibility equivalent 764 to LATIN SMALL LETTER F (U+0066) and LATIN SMALL LETTER I 765 (U+0069). 767 Under SASLprep, the use of NFKC also handled the mapping of 768 fullwidth and halfwidth code points to their decomposition 769 mappings. 771 For migration purposes, operators might want to search their 772 database of usernames for names containing Unicode code points 773 with compatibility equivalents and, where there is no conflict, 774 map those code points to their equivalents. Naturally, it is 775 possible that during this process the operator will discover 776 conflicting usernames (e.g., HENRYIV with the last two code points 777 being LATIN CAPITAL LETTER I (U+0049) and LATIN CAPITAL LETTER V 778 (U+0056) vs. "HENRYIV" with the last character being ROMAN NUMERAL 779 FOUR (U+2163), which is compatibility equivalent to U+0049 and 780 U+0056); in these cases, the operator will need to determine how 781 to proceed -- for instance, by disabling the account whose name 782 contains a Unicode code point with a compatibility equivalent. 783 Such cases are probably rare, but it is important for operators to 784 be aware of them. 786 o SASLprep mapped the "characters commonly mapped to nothing" from 787 Appendix B.1 of [RFC3454]) to nothing, whereas the PRECIS 788 IdentifierClass entirely disallows most of these code points, 789 which correspond to the code points from the PRECIS "M" category 790 defined under Section 9.13 of [RFC7564]. For migration purposes, 791 the operator might want to remove from usernames any code points 792 contained in the PRECIS "M" category (e.g., SOFT HYPHEN (U+00AD)). 793 Because these code points would have been "mapped to nothing" in 794 stringprep, in practice a user would not notice the difference if, 795 upon migration to PRECIS, the code points are removed. 797 o SASLprep allowed uppercase and titlecase code points, whereas the 798 UsernameCaseMapped profile maps uppercase and titlecase code 799 points to their lowercase equivalents (by contrast, the 800 UsernameCasePreserved profile matches SASLprep in this regard). 801 For migration purposes, the operator can use either the 802 UsernameCaseMapped profile (thus losing the case information) or 803 the UsernameCasePreserved profile (thus ignoring case difference 804 when comparing usernames). 806 6.2. Passwords 808 Depending on local service policy, migration from RFC 4013 to this 809 specification might not involve any scrubbing of data (because 810 passwords might not be stored in the clear anyway); however, service 811 providers need to be aware of possible issues that might arise during 812 migration. In particular: 814 o SASLprep specified the use of Unicode Normalization Form KC 815 (NFKC), whereas the OpaqueString profile employs Unicode 816 Normalization Form C (NFC). Because NFKC is more aggressive about 817 finding matches than NFC, in practice this change is unlikely to 818 cause significant problems and indeed has the security benefit of 819 probably resulting in fewer false positives when comparing 820 passwords. A few examples might suffice to indicate the nature of 821 the problem: 823 1. LATIN SMALL LETTER LONG S (U+017F) is compatibility equivalent 824 to LATIN SMALL LETTER S (U+0073). 826 2. ROMAN NUMERAL FOUR (U+2163) is compatibility equivalent to 827 LATIN CAPITAL LETTER I (U+0049) and LATIN CAPITAL LETTER V 828 (U+0056). 830 3. LATIN SMALL LIGATURE FI (U+FB01) is compatibility equivalent 831 to LATIN SMALL LETTER F (U+0066) and LATIN SMALL LETTER I 832 (U+0069). 834 Under SASLprep, the use of NFKC also handled the mapping of 835 fullwidth and halfwidth code points to their decomposition 836 mappings. Although it is expected that code points with 837 compatibility equivalents are rare in existing passwords, some 838 passwords that matched when SASLprep was used might no longer work 839 when the rules in this specification are applied. 841 o SASLprep mapped the "characters commonly mapped to nothing" from 842 Appendix B.1 of [RFC3454]) to nothing, whereas the PRECIS 843 FreeformClass entirely disallows such code points, which 844 correspond to the code points from the PRECIS "M" category defined 845 under Section 9.13 of [RFC7564]. In practice, this change will 846 probably have no effect on comparison, but user-oriented software 847 might reject such code points instead of ignoring them during 848 password preparation. 850 7. IANA Considerations 852 IANA has made the updates described below. 854 7.1. UsernameCaseMapped Profile 856 IANA has added the following entry to the "PRECIS Profiles" registry. 858 Name: UsernameCaseMapped. 860 Base Class: IdentifierClass. 862 Applicability: Usernames in security and application protocols. 864 Replaces: The SASLprep profile of stringprep. 866 Width-Mapping Rule: Map fullwidth and halfwidth code points to their 867 decomposition mappings. 869 Additional Mapping Rule: None. 871 Case-Mapping Rule: Map uppercase and titlecase code points to 872 lowercase. 874 Normalization Rule: NFC. 876 Directionality Rule: The "Bidi Rule" defined in RFC 5893 applies. 878 Enforcement: To be defined by security or application protocols that 879 use this profile. 881 Specification: RFC 7613 (this document), Section 3.2. 883 7.2. UsernameCasePreserved Profile 885 IANA has added the following entry to the "PRECIS Profiles" registry. 887 Name: UsernameCasePreserved. 889 Base Class: IdentifierClass. 891 Applicability: Usernames in security and application protocols. 893 Replaces: The SASLprep profile of stringprep. 895 Width-Mapping Rule: Map fullwidth and halfwidth code points to their 896 decomposition mappings. 898 Additional Mapping Rule: None. 900 Case-Mapping Rule: None. 902 Normalization Rule: NFC. 904 Directionality Rule: The "Bidi Rule" defined in RFC 5893 applies. 906 Enforcement: To be defined by security or application protocols that 907 use this profile. 909 Specification: RFC 7613 (this document), Section 3.3. 911 7.3. OpaqueString Profile 913 IANA has added the following entry to the "PRECIS Profiles" registry. 915 Name: OpaqueString. 917 Base Class: FreeformClass. 919 Applicability: Passwords and other opaque strings in security and 920 application protocols. 922 Replaces: The SASLprep profile of stringprep. 924 Width-Mapping Rule: None. 926 Additional Mapping Rule: Map non-ASCII space code points to ASCII 927 space. 929 Case-Mapping Rule: None. 931 Normalization Rule: NFC. 933 Directionality Rule: None. 935 Enforcement: To be defined by security or application protocols that 936 use this profile. 938 Specification: RFC 7613 (this document), Section 4.2. 940 7.4. Stringprep Profile 942 The stringprep specification [RFC3454] did not provide for entries in 943 the "Stringprep Profiles" registry to have any state except "Current" 944 or "Not Current". Because this document obsoletes RFC 4013, which 945 registered the SASLprep profile of stringprep, IANA has marked that 946 profile as "Not Current" and cited this document as an additional 947 reference. 949 8. Security Considerations 951 8.1. Password/Passphrase Strength 953 The ability to include a wide range of characters in passwords and 954 passphrases can increase the potential for creating a strong password 955 with high entropy. However, in practice, the ability to include such 956 characters ought to be weighed against the possible need to reproduce 957 them on various devices using various input methods. 959 8.2. Identifier Comparison 961 The process of comparing identifiers (such as SASL simple user names, 962 authentication identifiers, and authorization identifiers) can lead 963 to either false negatives or false positives, both of which have 964 security implications. A more detailed discussion can be found in 965 [RFC6943]. 967 8.3. Reuse of PRECIS 969 The security considerations described in [RFC7564] apply to the 970 IdentifierClass and FreeformClass base string classes used in this 971 document for usernames and passwords, respectively. 973 8.4. Reuse of Unicode 975 The security considerations described in [UTS39] apply to the use of 976 Unicode code points in usernames and passwords. 978 9. References 980 9.1. Normative References 982 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 983 Requirement Levels", BCP 14, RFC 2119, 984 DOI 10.17487/RFC2119, March 1997, 985 . 987 [RFC3629] Yergeau, F., "UTF-8, a transformation format of ISO 988 10646", STD 63, RFC 3629, DOI 10.17487/RFC3629, November 989 2003, . 991 [RFC5234] Crocker, D., Ed. and P. Overell, "Augmented BNF for Syntax 992 Specifications: ABNF", STD 68, RFC 5234, 993 DOI 10.17487/RFC5234, January 2008, 994 . 996 [RFC5890] Klensin, J., "Internationalized Domain Names for 997 Applications (IDNA): Definitions and Document Framework", 998 RFC 5890, DOI 10.17487/RFC5890, August 2010, 999 . 1001 [RFC6365] Hoffman, P. and J. Klensin, "Terminology Used in 1002 Internationalization in the IETF", BCP 166, RFC 6365, 1003 DOI 10.17487/RFC6365, September 2011, 1004 . 1006 [RFC7564] Saint-Andre, P. and M. Blanchet, "PRECIS Framework: 1007 Preparation, Enforcement, and Comparison of 1008 Internationalized Strings in Application Protocols", 1009 RFC 7564, DOI 10.17487/RFC7564, May 2015, 1010 . 1012 [UAX11] Unicode Standard Annex #11, "East Asian Width", edited by 1013 Ken Lunde. An integral part of The Unicode Standard, 1014 . 1016 [Unicode] The Unicode Consortium, "The Unicode Standard", 1017 . 1019 9.2. Informative References 1021 [Err1812] RFC Errata, "Erratum ID 1812", RFC 4013, 1022 . 1024 [RFC20] Cerf, V., "ASCII format for network interchange", STD 80, 1025 RFC 20, DOI 10.17487/RFC0020, October 1969, 1026 . 1028 [RFC3454] Hoffman, P. and M. Blanchet, "Preparation of 1029 Internationalized Strings ("stringprep")", RFC 3454, 1030 DOI 10.17487/RFC3454, December 2002, 1031 . 1033 [RFC3501] Crispin, M., "INTERNET MESSAGE ACCESS PROTOCOL - VERSION 1034 4rev1", RFC 3501, DOI 10.17487/RFC3501, March 2003, 1035 . 1037 [RFC4013] Zeilenga, K., "SASLprep: Stringprep Profile for User Names 1038 and Passwords", RFC 4013, DOI 10.17487/RFC4013, February 1039 2005, . 1041 [RFC4422] Melnikov, A., Ed. and K. Zeilenga, Ed., "Simple 1042 Authentication and Security Layer (SASL)", RFC 4422, 1043 DOI 10.17487/RFC4422, June 2006, 1044 . 1046 [RFC4616] Zeilenga, K., Ed., "The PLAIN Simple Authentication and 1047 Security Layer (SASL) Mechanism", RFC 4616, 1048 DOI 10.17487/RFC4616, August 2006, 1049 . 1051 [RFC5802] Newman, C., Menon-Sen, A., Melnikov, A., and N. Williams, 1052 "Salted Challenge Response Authentication Mechanism 1053 (SCRAM) SASL and GSS-API Mechanisms", RFC 5802, 1054 DOI 10.17487/RFC5802, July 2010, 1055 . 1057 [RFC5893] Alvestrand, H., Ed. and C. Karp, "Right-to-Left Scripts 1058 for Internationalized Domain Names for Applications 1059 (IDNA)", RFC 5893, DOI 10.17487/RFC5893, August 2010, 1060 . 1062 [RFC6120] Saint-Andre, P., "Extensible Messaging and Presence 1063 Protocol (XMPP): Core", RFC 6120, DOI 10.17487/RFC6120, 1064 March 2011, . 1066 [RFC6943] Thaler, D., Ed., "Issues in Identifier Comparison for 1067 Security Purposes", RFC 6943, DOI 10.17487/RFC6943, May 1068 2013, . 1070 [RFC7542] DeKok, A., "The Network Access Identifier", RFC 7542, 1071 DOI 10.17487/RFC7542, May 2015, 1072 . 1074 [RFC7613] Saint-Andre, P. and A. Melnikov, "Preparation, 1075 Enforcement, and Comparison of Internationalized Strings 1076 Representing Usernames and Passwords", RFC 7613, 1077 DOI 10.17487/RFC7613, August 2015, 1078 . 1080 [RFC7616] Shekh-Yusef, R., Ed., Ahrens, D., and S. Bremer, "HTTP 1081 Digest Access Authentication", RFC 7616, 1082 DOI 10.17487/RFC7616, September 2015, 1083 . 1085 [RFC7617] Reschke, J., "The 'Basic' HTTP Authentication Scheme", 1086 RFC 7617, DOI 10.17487/RFC7617, September 2015, 1087 . 1089 [RFC7622] Saint-Andre, P., "Extensible Messaging and Presence 1090 Protocol (XMPP): Address Format", RFC 7622, 1091 DOI 10.17487/RFC7622, September 2015, 1092 . 1094 [UTS39] Unicode Technical Standard #39, "Unicode Security 1095 Mechanisms", edited by Mark Davis and Michel Suignard, 1096 . 1098 Appendix A. Changes from RFC 7613 1100 The following changes were made from [RFC7613]. 1102 o Corrected the order of operations for the UsernameCaseMapped 1103 profile to ensure consistency with RFC 7564. 1105 o In accordance with working group discussions and updates to 1106 [RFC7564], removed the use of the Unicode CaseFold() operation in 1107 favor of the Unicode toLower() operation. 1109 o Modified the presentation (but not the content) of the rules. 1111 o Removed UTF-8 as a mandatory encoding, because that is a matter 1112 for the application. 1114 o Clarified several editorial matters. 1116 o Updated references. 1118 See [RFC7613] for a description of the differences from [RFC4013]. 1120 Appendix B. Acknowledgements 1122 Thanks to Christian Schudt and Sam Whited for their bug reports and 1123 feedback. 1125 See [RFC7613] for acknowledgements related to the specification that 1126 this document supersedes. 1128 Authors' Addresses 1130 Peter Saint-Andre 1131 Filament 1132 P.O. Box 787 1133 Parker, CO 80134 1134 USA 1136 Phone: +1 720 256 6756 1137 Email: peter@filament.com 1138 URI: https://filament.com/ 1140 Alexey Melnikov 1141 Isode Ltd 1142 5 Castle Business Village 1143 36 Station Road 1144 Hampton, Middlesex TW12 2BX 1145 United Kingdom 1147 Email: Alexey.Melnikov@isode.com